Evaluation of Escherichia coli DJ4309 expressing human P450 1A2 in mutagenicity testing of complex food mixtures

Mutation Research 442 Ž1999. 79–87 www.elsevier.comrlocatergentox Community address: www.elsevier.comrlocatermutres

Evaluation of Escherichia coli DJ4309 expressing human P450 1A2 in mutagenicity testing of complex food mixtures Anne Constable

a,)

, Natacha Varga a , P. David Josephy b, Phillipe Guy a , Robert J. Turesky a

a

b

Nestle´ Research Center, Nestec, Vers-chez-les-Blanc, 1000 Lausanne 26, Switzerland Guelph-Waterloo Centre for Graduate Work in Chemistry, Department of Chemistry and Biochemistry, UniÕersity of Guelph, Guelph, Ontario, Canada N1G 2W1 Received 2 February 1998; received in revised form 14 April 1999; accepted 20 April 1999

Abstract Heterocyclic aromatic amines ŽHAAs. are potent bacterial mutagens and potential human carcinogens formed in heat processed proteins. The Ames test Žstrain TA98. is a useful mutagenicity test system to screen food products for these compounds. HAAs require activation to their genotoxic forms, and in the Ames test, a rat liver S-9 preparation is normally used. In order to better understand the mechanisms of mutagen activation with respect to human metabolism, new bacterial strains containing human cytochrome P450s and other metabolic enzymes have recently been developed. We have investigated the capacity of one of these strains, DJ4309 wJosephy et al., Chem. Res. Toxicol. 11 Ž1998. 70–74x as a screening tool for mutagens in food products. DJ4309 expresses the human P450 1A2, human NADPH cytochrome reductase and the bacterial acetyl CoA:arylamine N-acetyltransferase. This strain is as sensitive as the Ames system to the mutagenic effects of the heterocyclic aromatic amines 2-amino-3-methylimidazow4,5-f xquinoline, 2-amino-3,4-dimethylimidazow4,5-f xquinoline and 2-amino-3,8-dimethylimidazow4,5-f xquinoxaline, but less sensitive to 2-amino-1-methyl-6-phenylimidazow4,5-b xpyridine. However, the mutagenicity of the arylamine 2-aminofluorene is considerably higher in DJ4309 than in the Ames test system. Meat extracts with a total HAA content ranging from less than 2 ngrg to 20 ngrg are efficiently detected by the Ames TA98 strain with rat liver S-9 activation. DJ4309 is less sensitive, with fewer revertants induced over the same dose range. Unknown compounds present in the meat extracts appear to inhibit the activity of the P450 1A2 enzyme in the DJ4309 strain. We have therefore demonstrated that although DJ4309 is a useful tool for mechanistic studies in chemical carcinogenesis, the screening of complex food matrices for HAAs by this bacterial strain must be conducted with caution. q 1999 Elsevier Science B.V. All rights reserved. Keywords: Heterocyclic aromatic amine; Meat extract; DJ4309; Ames test

AbbreÕiations: HAAs, Heterocyclic aromatic amines; IQ, 2-amino-3-methylimidazow4,5-f xquinoline; MeIQ, 2-amino-3,4-dimethylimidazow4,5-f xquinoline; MeIQx, 2-amino-3,8-dimethylimidazow4,5-f xquinoxaline; 4,8-DiMeIQx, 2-amino-3,4,8-trimethylimidazow4,5f xquinoxaline; 7,8-DiMeIQx, 2-amino-3,7,8-trimethylimidazow4,5-f xquinoxaline; PhIP, 2-amino-1-methyl-6-phenylimidazow4,5-b xpyridine; 2-AF, 2-aminofluorene; DMSO, dimethylsulphoxide; d-ALA, d-aminolevulinic acid; IPTG, isopropyl b-D-thiogalactose; LB, Luria Broth ) Corresponding author. Tel.: q41-21-785-8574; Fax: q41-21-785-8553; E-mail: [email protected] 1383-5718r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved. PII: S 1 3 8 3 - 5 7 1 8 Ž 9 9 . 0 0 0 6 6 - 2

80

A. Constable et al.r Mutation Research 442 (1999) 79–87

1. Introduction Heterocyclic aromatic amines w1x are formed through Maillard reactions, the same chemistry which is responsible for the flavours and roasted aromas that develop during the cooking and roasting of meat. Amino acids, sugars, and creatine derived from muscle are the main precursors for both aroma production and the formation of undesirable mutagenic HAAs in high temperature reactions w2x. The predominant HAAs in cooked meats are IQ, MeIQ, MeIQx, 4,8-DiMeIQx, 7,8-DiMeIQx and PhIP ŽFig. 1.. After metabolic activation, HAAs can readily bind covalently to cellular DNA, inducing mutations which may ultimately lead to abnormal biochemical processes. Animal feeding studies using rodents and non-human primates indicate that these compounds produce malignant tumours at multiple sites w3x. Human tissues also activate HAAs at appreciable levels and are therefore suspect human carcinogens w4x. HAAs are potent bacterial mutagens, and the Salmonella typhimurium Ames tester strains TA1538 and TA98 are extremely sensitive to the mutagenic effects of activated HAAs w5x. The Ames test combined with chemical analysis is the most effective method for screening food products for HAAs. However, the source of metabolic activation in the Ames test is generally the S-9 fraction of rat liver homogenates, which may have different patterns of drug and carcinogen metabolism as compared to humans, weakening the validity of human health risk

estimates based on short term in-vitro assays. Mechanistically, the biochemistry of S-9 dependent metabolism is complex, and many enzyme activities, both activating and detoxifying, are present in the liver homogenate. Over recent years, new bacterial strains and eucaryotic cell lines have been engineered with the aim of replacing the requirement of rat hepatic S-9 for the activation of chemical mutagens w6x. Mutagenic aromatic and heterocyclic aromatic amines are mainly activated by P450 1A2 via N-hydroxylation. The resulting N-hydroxyarylamines can be further activated by phase II enzymes such as N-acetyltransferase ŽNAT. catalysed O-acetylation, yielding reactive N-acetoxy esters w7x. An Escherichia coli strain ŽDJ4309, Ref. w8x. has been constructed which contains the human P450 1A2 and NADPH reductase, together with overexpressed bacterial O-acetyl transferase. The objective of this study was to evaluate the efficiency of strain DJ4309 in mutagenicity screening tests of food products with respect to HAAs, and to compare biological activities to the existing Ames test system.

2. Materials and methods 2.1. Chemicals IQ, MeIQ, MeIQx and PhIP were purchased from Toronto Research Chemicals ŽDownsview, Ontario,

Fig. 1. Structure of HAAs commonly found in cooked meats.

A. Constable et al.r Mutation Research 442 (1999) 79–87

Canada.; 2-AF, d-ALA and IPTG were purchased from Sigma. Concentrated stock solutions of mutagens were prepared in DMSO and stored at y208C. NHOH-IQ and NHOH-PhIP were synthesised as described previously w9x, and HPLC analysis revealed that both derivatives were greater than 95% pure. Deuterated internal standards w2 H 3 x4,8-DiMeIQx, w2 H 3 x7,8-DiMeIQx and w2 H 3 xIQ were synthesised as previously described w10x and w2 H 3 xMeIQx was purchased from Toronto Research Chemicals ŽDownsview, Ontario, Canada.; w2 H 5 xPhIP was a kind gift from Drs. J. Felton and M. Knize, Lawrence Livermore National Laboratory, CA, USA. Meat extracts labelled A through D are commercially available. 2.2. Tester strains The E. coli strain DJ4309 was described by Josephy et al. w8x. DJ4309 is based on the E. coli strain CC109 which carries an FX factor bearing a q2 frameshift mutation in the sequence GCGCGCGC of the lacZ gene w11x. This is similar to the hisD3052 hotspot site found in S. typhimurium Ames tester strains TA1538 and TA98. A null mutation has been introduced into the uÕrA gene w12x rendering the bacteria defective in DNA nucleotide excision repair and therefore increasing the sensitivity to the effects of aromatic amine mutagenesis. The final strain, DJ4309, carries two compatible plasmids. Plasmid pNM12 carries the S. typhimurium NAT gene w13x which is more efficient in HAA metabolism than either of the two human enzymes, NAT1 or NAT2 w14x. The plasmid pCW1A2 reductase carries the genes for the human P450 1A2 and the human NADPH-cytochrome P450 reductase, both under the control of the IPTG inducible tac promotor. This strain is therefore metabolically competent to activate aromatic amine mutagens, resulting in revertant E. coli cells able to grow on lactose minimal media. The Ames tester strain TA98, kindly provided by Prof. B. Ames, University of California at Berkeley, USA, is the standard strain used in our laboratories for routine testing of foods for the presence of mutagenic HAAs. 2.3. Purification of meat extracts for mutagenicity testing The solid phase extraction method specific for the isolation of HAAs from complex mixtures as de-

81

scribed in Ref. w15x was used, with some modifications. Briefly, each meat extract was homogenised in 1 N NaOH prior to serial extraction with dichloromethane through Extrelut-20 ŽMerck. and PRS Žpropyl sulfonic acid silica, Isolute International Sorbent Technology. cartridges. The PRS cartridge was then eluted with 1 M ammonium acetate pH 8 onto a C-18 Sep Pak ŽIsolute C-18 MF, Isolute International Sorbent Technology.. Partially purified HAAs were eluted from the C-18 cartridge by methanol:ammonia Ž9:1.. Alternatively, the HAAs were further purified into polar ŽPhIP. and apolar ŽIQ type. HAAs. After the initial dichloromethane clean up through Extrelut and PRS, the PRS cartridge was eluted with methanol:1 N HCl Ž6:4.. Concentrated ammonia was added to this eluate to obtain a final pH of 10, applied to a C-18 cartridge, and the apolar HAAs eluted with methanol:ammonia Ž9:1.. The same PRS cartridge was then washed through with 1 M ammonium acetate pH 8 onto another C-18 Sep Pak, and the polar HAAs eluted with methanol:ammonia Ž9:1.. All extracts were dried, resuspended in an appropriate volume of DMSO and stored at y808C until tested. 2.4. Mutagenicity tests 2.4.1. DJ4309 Frozen stocks were prepared as described w8x and stored at y808C. An aliquot was rapidly thawed and 50 ml used to inoculate 50 ml LB medium supplemented with 50 mgrml ampicillin, 10 mgrml chloramphenicol, 100 mM IPTG, 0.5 mM d-ALA and 12.5 ml trace element mix w16x. It was observed that for reproducible bacterial growth, spontaneous rate and mutagenic response, this mix should be prepared fresh for each experiment and not stored. The optimum concentration of IPTG for mutagen activation was determined to be 100 mM Ždata not shown., and not 1 mM as previously used w8x. In addition, 50 ml overnight cultures were prepared instead of 5 ml cultures. The cultures were incubated at 308C, 150 rpm for exactly 12 h and thereafter kept on ice. The mutagenicity test was performed by mixing 0.5 ml sodium phosphate buffer Ž0.1 M, pH 7.4., appropriate quantities of mutagen, meat extract, or DMSO alone, in a 10 ml volume, and 0.1 ml bacterial culture. Pre-incubation was carried out at room tem-

82

A. Constable et al.r Mutation Research 442 (1999) 79–87

perature for 30 min with occasional shaking. After the addition of 2 ml top agar, the mixes were poured onto minimal lactose plates and incubated at 378C for 48 h. 2.5. Ames test The standard Ames direct plating test w17x was used. The tests were mixed with 0.1 ml of an overnight culture of S. typhimurium TA98 and 0.5 ml of freshly prepared S-9 mix ŽCo-factors and 10% S-9, prepared from the livers of male Sprague Dawley rats pretreated with Aroclor 1254 and obtained from Moltox, Annapolis, USA., added to 2 ml soft agar and plated on minimal Ž-histidine. agar plates. The Petri plates were incubated for 3 days at 378C. Colonies were counted on an automatic colony counter, Fisher Count-All 880. Individual experiments were performed at least twice, with triplicate platings at each dose. Average values are shown in Results. Spontaneous numbers of DJ4309 revertants ranged from 50 to 250; spontaneous numbers of TA98 revertants with S-9 ranged from 30 to 70. 2.6. HAA quantification An aliquot of 4 g of each sample was homogenised with 15–20 ml 1 N NaOH and 40 ml of a 500 ngrml mix of internal deuterated standards added to have a final internal standard concentration of 5 ngrg. The purification of apolar and polar HAAs was performed as described w15x, where HAAs were quantified by LC-APCIMSrMS using a Hewlett-Packard 1100 HPLC system coupled to a Finnigan TSQ-7000 mass spectrometer w18x.

3. Results and discussion 3.1. Analysis of pure HAAs A pre-incubation step was required for activation of HAAs by DJ4309 ŽFig. 2.. A pre-incubation did not, however, change the mutagenic response by TA98 plus rat liver S-9, and so this step was not included in the subsequent experiments for TA98. The mutagenic responses of TA98 with rat liver S-9

Fig. 2. Effect of pre-incubation on the mutagenicity of IQ. TA98, squares; DJ4309, triangles. Open symbols indicate pre-incubation.

and DJ4309 to individual HAAs are shown in Fig. 3. For comparison, the number of revertants induced per ng of test compound was calculated from the slope of the linear part of the dose response curves ŽTable 1.. None of the HAAs are mutagenic in TA98 without S9, or in strain DJ3109-pNM12 which does not express P450 1A2 Ždata not shown.. IQ, MeIQ and MeIQx gave similar responses from both strains, in that ng amounts of the compounds tested induced mutagenic responses. It has been previously demonstrated that preincubation of DJ4309 with a-naphthoflavone w8x inhibits the mutagenicity of these compounds, indicating the requirement of P450 1A2 in this system. The mutagenic potencies of all the mutagens tested in TA98 with S-9 were, in decreasing order, MeIQ) IQ ) MeIQx) PhIP ) 2-AF. For DJ4309, the mutagenic potencies were MeIQrIQ) MeIQx) 2-AF ) PhIP. The most potent HAA in both strains is MeIQ. IQ is a weaker mutagen in the Ames test, but DJ4309 showed a similar response to IQ as to MeIQ. PhIP is the weakest mutagenic HAA tested and microgram amounts are required to induce similar responses to the IQ compounds in TA98. DJ4309 shows a very weak response and up to 10 mg of PhIP was required to reproducibly increase the number of revertants over background levels. The response of DJ4309 to the HAAs is consistent with previous results w8x. Considerable differences were observed in sensitivity of the two strains to the arylamine 2-AF. In the DJ4309 strain, 93,000 revertants were induced per mg of 2-AF, whereas at the same dose only 236 revertants of TA98 are observed ŽFig. 4 and Table 2..

A. Constable et al.r Mutation Research 442 (1999) 79–87

83

Fig. 3. Comparison of the mutagenic effects of MeIQ, MeIQx, IQ and NHOH-IQ, and PhIP and NHOH-PhIP in TA98 with rat liver S-9 mediated activation and in DJ4309 with human P450 1A2 activation. TA98, shaded triangles, DJ4309, open triangles. NHOH-IQ and NHOH-PhIP was tested with TA98 in the absence of S-9 Žshaded boxes with dashed line.. NHOH-IQ and NHOH-PhIP with DJ4309, open boxes with dashed line.

It has previously been shown that the catalytic efficiency and N-oxidation rates of MeIQx and PhIP were greater with human liver microsomes and purified recombinant human P450 1A2, than with rat liver microsomes and purified rat P450 1A2 w19x. These species differences in P450 1A2 activity, together with the overexpression of the bacterial OAT, may account for some of the different activities of Table 1 Mutagenic potency of mutagens in TA98 with S-9 and in DJ4309 TA98

MeIQx, IQ, PhIP and 2-AF observed between the two systems in this study. Quantitative differences in P450 1A2 activity would also be a contributing factor; these were not measured. P450 1A2 is not the sole P450 to metabolise HAAs. PCB-pretreated rat liver S-9 also contains significant amounts of P450 1A1 which is also capable of N-hydroxylating HAAs w20x. Numerous other metabolic enzymes are also present in the S-9 preparation ŽP450s, glutathione S-transferases, UDP-glucuronyl transferases and

DJ4309

Revertantsrng MeIQ IQ NHOH-IQ MeIQx

666 157 1798 70

622 642 10,580 174

Revertantsrmg PhIP NHOH-PhlP 2-AF

704 5530 236

32 424 93,000

The values are taken from the slope of regression on the linear part of the dose response curves and indicate the number of revertants induced per ng ŽMeIQ, IQ, NHOH-IQ, MeIQx. or per mg ŽPhIP, NHOH-PhIP, 2-AF. of mutagen.

Fig. 4. Mutagenicity of 2-aminofluorene. TA98, shaded triangles, DJ4309, open triangles.

A. Constable et al.r Mutation Research 442 (1999) 79–87

84

Table 2 Mutagenic potency of meat extracts in TA98 with S-9 and in DJ4309 Sample

A B C D

Meat extracts Žrevertantsrg E.

Polar extracts Žrevertantsrg E.

TA98

DJ4309

DJ4309

1950 1739 751 241

180 221 40 19

280 227 94 21

Values indicate the number of revertants induced per g E of meat extract.

sulphotransferases. which may contribute to a modulation of mutagenic activities. In order to address some of these points, the NHOH-IQ and NHOH-PhIP metabolites of IQ and PhIP were tested in the two bacterial systems. NHOH-IQ and NHOH-PhIP are formed from IQ and PhIP after N-hydroxylation by P450 1A2. Since the parent compounds are not necessarily converted quantitatively to their N-hydroxy derivatives, differences in potencies would be expected. It was observed that in contrast to PhIP and IQ, a pre-incubation step increased the mutagenic response of TA98 to both NHOH-IQ and NHOH-PhIP in the absence of S-9 Ždata not shown.. With TA98, 1,798 revertants per ng of NHOH-IQ were induced Žno S-9 added., compared to only 157 revertants per ng IQ Žwith S-9.. With DJ4309, 10,580 revertants per ng NHOH-IQ were induced compared to 642 revertants per ng IQ. This would suggest that the rat liver S-9 converts approximately 10% of the IQ to NHOH-IQ, while DJ4309 converts only 6%. However, the response in DJ4309 was actually 6 fold stronger than in TA98. The second step in activation of IQ to its mutagenic from is by the conversion to its N-acetoxy ester by O-acetyltransferase. The increased sensitivity of DJ4309 to both NHOH-IQ and IQ may be explained by the overexpression of the bacterial Oacetyltransferase on plasmid pN12, whereas TA98 has only one chromosomal copy. The greater sensitivity of the TA98 strain to the mutagenic effects of PhIP were again observed when NHOH-PhIP was tested. NHOH-PhIP induced 5530 revertants per mg with TA98 in the absence of S-9. Only 424 revertants of DJ4309 per mg NHOH-PhIP

were induced. The differences in sensitivity observed between the two strains cannot therefore be accounted for by differences of rat and human P450 1A2 specificity for PhIP, since the hydroxylation step is no longer a requirement with NHOH-PhIP. In fact, either the O-acetylation of NHOH-PhIP appears much less important than for the activation of NHOH-IQ in DJ4309 w14x, or more interestingly, NHOH-PhIP is metabolised by the bacterial enzyme to form a higher proportion of N, N-acetylhydroxyamino-PhIP, which does not react with DNA, than the DNA reactive N 2-acetoxy-PhIP w21x. Overexpression of the bacterial O-acetyltransferase in this case may therefore result in detoxification of PhIP. Similar bacterial strains have been constructed by different workers. The E. coli strain BMX100 w22x reverts to L-arginine prototrophy via base substitution mutations and contains the human CYP1A2 fused to rat liver NADPH cytochrome P450 reductase. IQ was weakly mutagenic in this strain, inducing only 202 revrnmol. This may due to the limiting NAT activity in strain BMX100, and IQ may be less efficient in inducing arginine reversion at the argE locus. Suzuki et al. w23x have introduced the human CYP1A2, human NADPH-cytochrome reductase and the Salmonella O-acetyl transferase into the Ames tester strain S. typhimurium TA1538, forming the strain TA1538rARO. The sensitivity of this strain to the mutagenic effects of HAAs was increased to that of the parent strain with S-9. Differences in sensitivity between differently engineered bacterial strains may reflect the different sensitivities of the E. coli and Salmonella bacteria in HAA uptake, different sensitivities of the target gene to mutation and also differences in expression levels of the cloned metabolic enzymes. However, a proper comparison of the efficiency of different strains to detect mutagenic activity must be compared within the same laboratory. 3.2. Mutagen testing of meat extracts The Ames test can be used as a screening tool to detect mutagenic activity in food samples and under appropriate conditions can give an indication of the presence of mutagenic HAAs. Chemical analyses should be performed on positive samples for more

A. Constable et al.r Mutation Research 442 (1999) 79–87

precise identification and quantification of known HAAs. Extracts of four different meat extracts were tested with TA98 and S-9. The results are shown in Fig. 5a and Table 2. These are typical dose dependent responses. Chemical analysis of three of these samples confirms that the meat extracts do indeed contain HAAs ŽTable 3.. MeIQx is the most abundant and neither MeIQ or PhIP were detected. The mutagenic response does not correlate exactly with the total HAA content, because the spectrum of individual HAAs present in each sample is different. Even though sample A contains a total of 20 ngrg HAAs, 75% of this is MeIQx, which shows weaker

Fig. 5. Mutagenicity of meat extracts. A,B,C,D refer to four different meat extract samples. The preparation of extracts is described in the material and methods. Doses are expressed in mg E of starting material per plate.

85

Table 3 Quantities of HAAs in meat extracts Žngrg. Meat extract

IQ

MeIQx

7.8diMeIQx

4.8diMeIQx

PhIP

Total

A B D

1.2 2.33 0.04

14.47 4.39 0.95

0.33 0.24 0.12

4.45 1.2 0.41

- 0.19 - 0.20 - 0.04

20.45 8.16 1.52

mutagenicity in the bacterial mutagenicity systems than IQ. Sample B contains twice the amount of IQ than sample A, which is more potent than MeIQx, therefore the mutagenicity of samples A and B are similar in the Ames test, despite a 2.5 fold difference in total HAA content. Low amounts of 7,8-DiMeIQx and 4,8-DiMeIQx were detected; however, they were not tested as individual pure compounds. The Ames test is sensitive enough under these conditions to detect meat extracts containing low levels of IQ type HAAs Žless than 2 ngrg—sample D.. Since the DJ4309 strain containing the human P450 1A2 has comparable and even improved mutagenic responses to IQ type compounds as the standard Ames assay using rat liver S-9 metabolic activation, we tested the efficiency of this strain in detecting HAAs in complex food mixtures. The same meat extracts as above, and using the same dose range of up to 1 g E, were tested in the strain DJ4309. A positive control of 5 ng IQ was included separately in each experiment to confirm that the bacteria were metabolically competent. Dose dependent responses were observed, with similar relative mutagenic responses as with TA98, but at much reduced absolute levels of revertants ŽFig. 5b and Table 2.. This was not due to toxicity, since the DJ4309 bacterial titre remained at 7.4–7.8 = 10 7 cellsrml even with the addition of 1 g E meat extract A, or with addition of an equal volume Ž10 ml. of DMSO solvent. The meat extracts were then further purified into the polar and apolar fractions in an effort to increase the mutagenic response. Similar weak responses were observed when the polar extracts were tested with DJ4309 ŽFig. 5c and Table 2.. No mutagenic activity was observed in the apolar fractions Ždata not shown.. This is expected as only PhIP, which was not detected in the meat extracts, is recovered in these fractions w15x.

86

A. Constable et al.r Mutation Research 442 (1999) 79–87

Although individual HAAs are potent mutagens in both DJ4309 and TA98, meat extracts shown to contain significant amounts of HAAs are less mutagenic in the DJ4309 strain than in TA98 with S-9. Co-incubation control experiments were performed. TA98 with S-9 was incubated with 5 ng MeIQ, with and without the addition of 1gE of meat extract A. The number of revertants induced were similar in both cases Ž1921 " 143, compared to 2270 " 23 with meat extract.. However, the number of DJ4309 revertants induced by 5 ng of MeIQ or of IQ were reduced by over 50% in the presence of the same meat extract Ž2329 " 34 to 1084 " 13, and 1273 " 31 to 430 " 72, respectively.. Both the polar and apolar fractions tested from each meat extract sample A to D had similar effects Ždata not shown., indicating that the inhibiting activity had no correlation with the HAA content, and was not removed by the extraction procedures. A reduction in mutagenic response from DJ4309 was not however observed with 1 ng NHOH-IQ, which does not require P450 1A2 activation Ž2191 " 317 revertants compared to 2357 " 20 in the presence of meat extract.. These results suggest that the human P450 1A2 in DJ4309 may be particularly sensitive to the presence of other components in the complex food mixture, which may either compete for P450 1A2 activity or act as inhibitors. Heat altered derivatives of linoleic acids, for example, have been shown to modulate the mutagenicity of fried ground beef w24x.

4. Conclusion E. coli strain DJ4309 expressing human P450 1A2, NADPH reductase and bacterial O-acetyltransferase is efficient at detecting mutagenicity from individual HAAs, and would therefore be a useful tool in mechanistic studies. However, the screening of complex food mixtures for HAAs by this tester strain would require further investigation regarding the purification of the test sample to eliminate possible inhibitory substances. DJ4309 may even be applied to the identification of these P450 inhibitors, which may modulate the genotoxicity of HAAs in the diet.

References w1x T. Sugimura, M. Nagao, T. Kawachi, M. Honds, T. Yahage, Y. Seino, S. Sata, N. Matsukara, T. Matushima, A. Shirai, M. Sawamura, H. Matsumoto, Mutagens–carcinogens in foods, with special reference to highly mutagenic pyrolytic products in broiled foods. Origins of Human Cancer, Cold Spring Harbour Laboratory, 1977, pp. 1561–1577. w2x M. Jagerstad, A. Laser-Reutersward, R. Olsen, S. Grivas, T. ¨ Nyhammar, K. Olsen, A. Dahlquist, CreatinŽine. and Maillard reaction products as precursors of mutagenic compounds: effects of various amino acids, Food Chem. 12 Ž1983. 255–264. w3x I.C. Munro, E. Kennepohl, R.E. Erickson, P.S. Portoghese, B.M. Wagner, O.D. Easterday, C.H. Manley, Safety assessment of ingested heterocyclic amines: initial report, Regul. Toxicol. Pharmacol. 17 Ž1993. S1–S109. w4x K. Wakabayashi, Y. Totsuka, K. Fukutome, A. Oguri, H. Ushiyama, T. Sugimura, Human exposure to mutagenicr carcinogenic heterocyclic amines and comutagenic betacarbolines, Mutat. Res. 376 Ž1997. 253–259. w5x A. Brahms, C. De Meester, Mutagenic potency of heterocyclic amines towards Salmonella typhimurium, Mutat. Res. 280 Ž1992. 103–107. w6x F.P. Guengerich, E.M.J. Gillam, T. Shimada, New applications of bacterial systems to problems in toxicology, Crit. Rev. Toxicol. 26 Ž1996. 551–583. w7x H.U. Aeschbacher, R.J. Turesky, Mammalian cell mutagenicity and metabolism of heterocyclic aromatic amines, Mutat. Res. 259 Ž1991. 235–250. w8x P.D. Josephy, D.H. Evans, A. Parikh, F.P. Guengerich, Metabolic activation of aromatic amine mutagens by simultaneous expression of human cytochrome P450 1A2, NADPH-cytochrome P450 reductase, and N-acetyltransferase in Escherichia coli, Chem. Res. Toxicol. 11 Ž1998. 70–74. w9x R.J. Turesky, N.P. Lang, M.A. Butler, C.H. Teitel, F.F. Kadulbar, Metabolic activation of carcinogenic heterocyclic aromatic amines by human liver and colon, Carcinogenesis 12 Ž1991. 1839–1845. w10x R.J. Turesky, H. Bur, H.U. Aeschbacher, H. Milon, Analysis of mutagenic heterocyclic amines in cooked beef by high performance liquid chromatography in combination with mass spectroscopy, Food Chem. Toxicol. 26 Ž1988. 501–509. w11x C.G. Cupples, M. Cabrera, C. Cruz, J.H. Miller, A set of lacZ mutations in Escherichia coli that allow rapid detection of specific frameshift mutations, Genetics 125 Ž1990. 275– 280. w12x T.M. Marwood, D. Meyer, P.D. Josephy, Escherichia coli lacZ strains engineered for detection of frameshift mutations induced by aromatic amines and nitroaromatic compounds, Carcinogenesis 16 Ž1995. 2037–2043. w13x Y. Oda, E. Yamazaki, M. Watanabe, T. Nohmi, T. Shimada, Highly sensitive umu test system for the detection of mutagenic nitroarenes in Salmonella typhimurium NM3009 having high O-acetyltransferase and nitroreductase activities, Environ. Mol. Mutagen. 21 Ž1993. 357–364.

A. Constable et al.r Mutation Research 442 (1999) 79–87 w14x D. Wild, W. Feser, S. Michel, H.L. Lord, P.D. Josephy, Metabolic activation of heterocyclic aromatic amines catalysed by human arylamine N-acetyltransferase isoenzymes ŽNAT1 and NAT2. expressed in Salmonella typhimurium, Carcinogenesis 16 Ž1995. 643–648. w15x G.A. Gross, A. Gruter, Quantitation of mutagenicrcarcino¨ genic heterocyclic aromatic amines in food products, J. Chromatogr. 592 Ž1992. 271–278. w16x P. Sandhu, Z. Guo, T. Baba, M.V. Martin, R.H. Tukey, F.P. Guengerich, Expression of modified human cytochrome P450 1A2 in Escherichia coli: stabilization, purification, spectral characterization and catalytic activities of the enzyme, Arch. Biochem. Biophys. 306 Ž1994. 443–450. w17x D.M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity test, Mutat. Res. 113 Ž1998. 178–215. w18x P. Guy, E. Gremaud, N. Varga, A. Constable, R.J. Turesky, 1998. Quantification of heterocyclic aromatic amines in meat based foods using liquid chromatographyratmospheric pressure chemical ionization tandem mass spectrometry. 15th Symposium on Liquid ChromatographyrMass Spectrometry. Montreux, Nov. 11–13. w19x R.J. Turesky, A. Constable, J. Richoz, N. Varga, J. Markovic, M.V. Martin, F.P. Guengerich, Activation of heterocyclic aromatic amines by rat and human liver microsomes and by purified rat and human cytochrome P450 1A2, Chem. Res. Toxicol. 11 Ž1998. 925–936.

87

w20x Y. Yamazoe, M. Abu-Zeid, S. Mnabe, S. Toyama, R. Kato, Metabolic activation of a protein pyrolysate promutagen 2-amino-3,8-dimethylimidazow4,5-f x by rat liver microsomes and purified cytochrome P450, Carcinogenesis 9 Ž1988. 105–109. w21x H. Frandsen, S. Grivas, R. Andersson, L. Dragsted, J.C. Larsen, Reaction of the N 2-acetoxy derivative of 2-aminoX 1-methyl-6-phenylimidazow4,5-b xpyridine ŽPhIP. with 2 -deoxyguanosine and DNA. Synthesis and identification of N 2Ž2X-deoxyguanosin-8-yl.-PhIP, Carcinogenesis 13 Ž1992. 629–635. w22x M. Kranendonk, P. Mesquita, A. Laires, N.P.E. Vermuelen, J. Rueff, Expression of human cytochrome P450 1A2 in Escherichia coli: a system for biotransformation and genotoxicity studies of chemical carcinogenesis, Mutagenesis 13 Ž1998. 263–269. w23x A. Suzuki, H. Kushida, H. Iwata, M. Watanabe, T. Nohmi, K. Fujita, F.J. Gonzalez, T. Kamataki, Establishment of a Salmonella tester strain highly sensitive to mutagenic heterocyclic amines, Cancer Res. 58 Ž1998. 1833–1838. w24x Y.L. Ha, N.K. Grimm, M.W. Pariza, Anticarcinogens from fried ground beef: heat altered derivatives of linoleic acid, Carcinogenesis 8 Ž1987. 1881–1887.